Unmanned vehicles (UVs) are seeing increased industry use as improvements in fields such as artificial intelligence, battery life, and computational are made. As an example, companies such as Amazon® are increasingly using UVs such as drones to deliver packages. As a result, some companies will likely begin to utilize hundreds or thousands of UVs at once to provide services.
Control over UVs may be complicated, due in part to a need to balance autonomous control with manual control. One particular use for UVs is controlling a fleet of UVs simultaneously, where control becomes exponentially more complicated. Manual control of each and every UV may be undesirable due to, e.g., excessive labor costs, human error, and the like.
Many solutions for automated navigation of UVs utilize on-board computations, thereby requiring more expensive hardware on-board each UV for performing computations. These costs are exacerbated when multiple UVs (i.e., a fleet) are controlled. Further, the computations needed to successfully navigate to target locations often become increasingly complex as an UV approaches the target. Specifically, as an UV approaches a target, travel by the UV requires more precise movements to, e.g., arrive at the correct geographical coordinates, avoid near-the-ground obstacles, land safely, and the like.
As a result of the difficulty in precisely navigating when approaching a landing site, existing UV solutions often face challenges arriving at particular sub-locations of a landing location. For example, many existing UV solutions cannot successfully pilot the UV to, e.g., a particular room or floor of a building. This inability to pilot to particular sub-locations can result in meddling with the UV, thereby frustrating its intended goal. As an example, when a drone delivers a package to an apartment complex, the drone may land in a general zone outside of the complex, which leaves the drone vulnerable to theft by a person other than the intended recipient.
Additionally, some existing solutions for landing UVs require the landing site to have a pre-known landing mat or other guide marker to successfully land. Such solutions can face challenges when the guide marker is obfuscated (e.g., if a visual guide marker is visually blocked or if signals from a guide marker are blocked or otherwise subject to interference). Further, if the landing site does not have a suitable guide marker, the UV may be unable to successfully navigate to the landing site.
Further, some existing solutions provide automated detection systems utilized to avoid and navigate around obstacles. Such solutions may still face challenges when obstacles are small or otherwise difficult for sensors of the UV to detect.
It would therefore be advantageous to provide a solution that would overcome the deficiencies of the prior art.